Characterization of Variable Platform for Robust Sensor and Separation Nanocomposite Membranes 1 Olson , J. M. W. Y. W. R. 1Chemical & Materials Engineering, San José State University, 2SRI International Objectives 2 Blum , Materials The goal of this research project is to create and characterize nanocomposite membranes using nanoparticles andsiloxane-based. Siloxane-based polymers have already been already reported as selective membranes for CO2. The chosen nanoparticles are nano graphene platelets that were recently discovered and are considered a low cost replacement of carbon nanotubes for enhancing mechanical properties of nanocomposite materials as well as electrical conductivity. Such composite membranes hold potential to produce a viable gas/electrical sensing membrane. Hence the creation and characterization of such materials is of interest. Results The siloxanes selected for this study are polydimethylsiloxane (PDMS), a polymer with numerous commercial applications, and modified polyhydromethylsiloxane (polyhydridomethylsiloxane, PHMS), based on a polymer platform developed by SRI International with great chemical diversity. Both polymers possess high thermal resistance and high CO2 selectivity. As a result of their molecular composition, they differ in their polarity and cross-linking characteristics. The conventional PDMS is non-polar, lightly crosslinked, and elastomeric in nature. Replacing one methyl group with a different functional or crosslinking group (e.g. modified PHMS, PRMS) can change the affinity and degree of crosslinking. Theory There are several mechanisms proposed to explain why the addition of a dispersed phase within a polymer matrix would enhance the gas permeability, selectivity, and sensing abilities of the material: ● Nanogaps – the dispersed phase is not perfectly adhered to the matrix, resulting in nanogaps along the interface through which gas might travel ● Polymer chain packing disruption - increases the free volume through which gases as passed ● Chemical modification – chemical interactions between the dispersed phase and the diffusing material Transport of large molecules is diffusion limited, while transport of small molecules is absorption limited. Therefore, mechanisms which increase diffusion, while not affecting absorption, may increase selectivity and sensing of larger molecules. In order to create the most efficient sensitivity with the least amount of material, the path the gas molecules must travel should be as long as possible. This will also increase the selectivity. To achieve this, a tortuous path is created, as in the figure below, left. The best possible shape for the dispersed phase is thus a flake, oriented perpendicular to the net path of gas flow. To this end, pre-exfoliated graphene has been selected as the dispersed phase to be placed in various crosslinked siloxane matrices. Adapted from: B.Z. Jang and W.C. Huang, “Nano-scaled graphene plates,” US Patent 7071258 (October 21, 2002). The structure of graphene, above, is rather like an unrolled carbon nanotube. Graphene shares many of the properties of carbon nanotubes like high thermal and electrical conductivity and strength. In fact, its shape may make it superior to nanotubes for the studied application. Lee, D. Cho, and L.T. Drzal, “Real-time observation of the expansion behavior of intercalated graphite flake,” J. Mater. Sci., 40, 231-234 (2005). It is important that the dispersed phase be exfoliated, as illustrated at right. This will result in maximization of the surface area for possible chemical reaction. Hay, J.N. and S.J. Shaw (2000). Clay-Based Composites [Online]. Available at http://www.azom.com/details.asp?ArticleID=936 (accessed September 21, 2007). WWW Article. Processing The process begins with using borosilicate glass filters as the backbone of the membrane. 1 Chung Since the graphene flakes are larger than the pores in these filters, any composite placed on top of them will be graphite enriched as the polymer permeates the glass fibers. These filters were thus soaked in PRMS and cured. K. Kalaitzidou, H. Fukushima, and L.T. Drzal, “Mechanical properties and morphological characterization of exfoliated graphite-polypropylene nanocomposites,” Compos. Part A, 38, 1675-1682 (2007). Lee, D. Cho, and L.T. Drzal, “Real-time observation of the expansion behavior of intercalated graphite flake,” J. Mater. Sci., 40, 231-234 (2005). A mixture of solvent, liquid PRMS, and graphene was created. The graphene made up 5% of the volume of the composite (polymer and graphene component of the mixture). The process of producing nano-scale graphene plates begins with a polymeric precursor. This precursor is then carbonized (fully or partially). This results in an intercalated graphite flake that looks like that pictured top left. This single flake is made up of many, many layers of graphene (middle left). Next, chemicals and heat are added to exfoliate the layers. The flake will expand from a thickness of 80-100 μm to a thickness of up to 10 mm or more in three seconds. Fast digital photography by Lee et al. captured this rapid transformation (bottom left). The exfoliated graphite is then ball-milled (or otherwise crushed) until nano-sized graphene, as shown bottom far left, results. The graphene flakes will consist of, as in the transmission electron microscope image at right, a small number of layers with total thickness on the nano-scale. The graphene platelets used in this project have an average diameter of 35 μm, with a maximum of 100 μm. In thickness, at least 80% are less than 100 nm. Scanning Electron Microscopy (SEM) was preformed on the PRMS/graphene nanocomposites, the pre-coated filters, and the raw filters. The Image at right was taken of a raw glass filter. The filter is clearly highly porous, and the permeability of the composites can not be determined by this mechanical support. The image of the polymercoated filter, shown below, indicates uniform coating throughout. SEM on the composites was performed on the surface and on a cross-section of a specimen cut with scissors. The composite appeared very rough, as shown below right. The rough nature of the composite increases the chance that there are straight paths through the composite for the gas. Also, literature evidence suggests that platelets laying perpendicular to the gas flow will be most effective. Future processing will focus on reducing this roughness. Coated glass filter Nanocomposite membrane K. Kalaitzidou, H. Fukushima, and L.T. Drzal, “Mechanical properties and morphological characterization of exfoliated graphite-polypropylene nanocomposites,” Compos. Part A, 38, 1675-1682 (2007). The composite solution was applied to the pre-coated filters and allowed to cure. Cross-sectional views of a specimen (above) showed clear delineation between the composite and the supporting substrate. The composite forms a nice, thick coating. Individual graphene flakes could be observed, as in right. It appears to be the expected radius. These flakes are so thin that structure beneath may be seen in the image. Gas Measurements Future Work The next step is to perform gas transport measurements and compare the performance of the polymer-only filters with the composite membranes. This will reveal if the roughness indicated by the SEM is providing fast paths through the membranes. These gas measurements will first be done with nitrogen gas and argon gas. Later carbon dioxide will be used, and possibly other gases. At right is the test chamber that has been designed and constructed for this purpose with automated computer data acquisition. ● Additional membranes will be created using smaller amounts of graphene by volume. ● Additional membranes will also be made with modified PHMS and PDMS. ● Process refinements will focus on creating flatter surfaces with flakes aligned parallel to the substrate. Acknowledgments J. Olson is supported by Defense Microelectronics Activity Cooperative Agreement #H94003-07-2-0705. Scanning electron microscope images were possible thanks to National Science Foundation grant #0421562 for Major Research Instrumentation. Graphene donated by Angston Materials, LLC. Thanks to: Prof. E. Allen, Prof. M. McNeil, D. Hui, A. Micheals, T. Olson , N. Peters and D. Verbosky.